From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S267651AbUH3Kxn (ORCPT ); Mon, 30 Aug 2004 06:53:43 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S267683AbUH3Kxn (ORCPT ); Mon, 30 Aug 2004 06:53:43 -0400 Received: from holomorphy.com ([207.189.100.168]:13236 "EHLO holomorphy.com") by vger.kernel.org with ESMTP id S267651AbUH3Kxm (ORCPT ); Mon, 30 Aug 2004 06:53:42 -0400 Date: Mon, 30 Aug 2004 03:53:22 -0700 From: William Lee Irwin III To: Paulo Marques Cc: Albert Cahalan , Roger Luethi , linux-kernel mailing list , Paul Jackson Subject: Re: [BENCHMARK] nproc: netlink access to /proc information Message-ID: <20040830105322.GE5492@holomorphy.com> Mail-Followup-To: William Lee Irwin III , Paulo Marques , Albert Cahalan , Roger Luethi , linux-kernel mailing list , Paul Jackson References: <20040828195647.GP5492@holomorphy.com> <20040828201435.GB25523@k3.hellgate.ch> <20040829160542.GF5492@holomorphy.com> <20040829170247.GA9841@k3.hellgate.ch> <20040829172022.GL5492@holomorphy.com> <20040829175245.GA32117@k3.hellgate.ch> <20040829181627.GR5492@holomorphy.com> <20040829190050.GA31641@k3.hellgate.ch> <1093810645.434.6859.camel@cube> <4133020F.1060306@grupopie.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <4133020F.1060306@grupopie.com> Organization: The Domain of Holomorphy User-Agent: Mutt/1.5.6+20040722i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Albert Cahalan wrote: >> This is crummy. It's done for wchan, since that is so horribly >> expensive, but I'm not liking the larger race condition window. >> Remember that PIDs get reused. There isn't a generation counter >> or UUID that can be checked. On Mon, Aug 30, 2004 at 11:31:43AM +0100, Paulo Marques wrote: > I just wanted to call your attention to the kallsyms speedup patch that > is now on the -mm tree. > It should improve wchan speed. My benchmarks for kallsyms_lookup (the > function that was responsible for the wchan time) went from 1340us to 0.5us. > So maybe this is enough not to make wchan a special case anymore... This seems to go wrong on big-endian machines; any chance you could look over your stuff and try to figure out what endianness issues it may have? -- wli